Medical image compression using adaptive neural network

High efficiency image compression methods are under wide attention for storage and transmission of images because images have large data quantity. This paper proposes a back propagation neural network approach used for data compression which offers parallel processing of data, training process makes the network suitable for various kinds of data. The number of units in the input and output layers of neural network are same and the number units in the hidden layer are smaller than other layers. The network is trained to perform the identity mapping. The compressed image is in the output of the hidden layer and decompressed image is in the output of the output layer of neural network. The proposed method experimentally proves that it provide reconstructed image with good quality than existing methods.